Research map / Label-efficient and robust learning / OOD and anomaly detection
Old Is Gold: Redefining the Adversarially Learned One-Class Classifier Training Paradigm
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use the two-phase OGNet training framework with both generator and discriminator at test time | Adversarial one-class anomaly detection · UCSD Ped2 | frame-level AUC: 92.9% -> 98.1%; absolute gain of 5.2% | Helped |
| Use OGNet instead of the strongest reported baseline | Image novelty detection · Caltech-256, 1 inlier class | AUC: 94.8% -> 98.2% | Helped |
| Use OGNet instead of the strongest reported baseline | Image novelty detection · Caltech-256, 3 inlier classes | AUC: 93.8% -> 97.7% | Helped |
| Use OGNet instead of the strongest reported baseline | Image novelty detection · Caltech-256, 5 inlier classes | AUC: 92.3% -> 98.1% | Helped |
| Add real images and low-quality reconstructions from the old generator to phase two discriminator training (ablation) | OGNet · UCSD Ped2 | frame-level AUC: 92.9% -> 94.4% | Helped |
| Add real images X to phase-two discriminator training (ablation) | OGNet · UCSD Ped2 | frame-level AUC: 94.4% -> 95.1% | Helped |
| Add pseudo-anomaly reconstructions to phase-two discriminator training (ablation) | OGNet · UCSD Ped2 | frame-level AUC: 95.1% -> 98.1% | Helped |
| Add low-quality reconstructions from the old generator to phase-two discriminator training (ablation) | OGNet · UCSD Ped2 | frame-level AUC: 95.9% -> 98.1% | Helped |
| Replace regenerated pseudo-anomalies with raw blended fake anomalies (ablation) | OGNet · UCSD Ped2 | frame-level AUC: 9.6% drop in the AUC; 88.5% | Hurt |
| Choose different low-epoch generators, or average parameters of previous generators, as the old generator (ablation) | OGNet · UCSD Ped2 | AUC | No clear effect |
Closest papers by meaning
- Contamination-Resilient Anomaly Detection via Adversarial Learning on Partially-Observed Normal and Anomalous Data (ICML 2024 · similarity 0.89)
- FrameShield: Adversarially Robust Video Anomaly Detection (NeurIPS 2025 · similarity 0.87)
- Boosting Fine-Grained Visual Anomaly Detection with Coarse-Knowledge-Aware Adversarial Learning (AAAI 2025 · similarity 0.87)
- Long-Tailed Anomaly Detection with Learnable Class Names (CVPR 2024 · similarity 0.87)
- PatchGuard: Adversarially Robust Anomaly Detection and Localization through Vision Transformers and Pseudo Anomalies (CVPR 2025 · similarity 0.86)
- ADPretrain: Advancing Industrial Anomaly Detection via Anomaly Representation Pretraining (NeurIPS 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.